This invention relates to methods for synthesis of oligosaccharides, and in particular to methods for solid phase or combinatorial synthesis of oligosaccharides. The invention provides a novel linker-resin, linker-saccharide, or resin-linker-saccharide complex, which in one embodiment enables a saccharide residue to be linked to a soluble or insoluble polymeric support for use as a basis for solid-phase synthesis of oligosaccharides. In a second embodiment, the complex of the invention enables oligosaccharides to be linked to a solid polymeric support for use as an analytical reagent.
Oligosaccharides constitute a major class of bioactive polymers, implicated in biochemical processes (Lasky, 1992; Varki, 1993) as diverse as cellular differentiation, hormone-cell recognition and cell-cell adhesion, especially viral-host cell (Gambaryan et al, 1995) and bacteria-host cell attachment (Boren et al, 1993). Involvement of oligosaccharides in diseases such as cancer, cardiovascular disorders, microbial infections, graft rejection and autoimmune disorders has therefore, been strongly suggested. Conjugation of carbohydrates to bioactive peptides has also been demonstrated to stabilise the peptides against degradation, and, in more specific circumstances, to facilitate peptide transport across biological barriers (Lee, 1989; Fisher, 1991; Rodriguez, 1989). Thus the ability to synthesise oligosaccharides in a facile and efficient manner is now becoming an extremely important area within organic chemistry.
The highly labour intensive solution phase strategies hitherto utilised in oligosaccharide syntheses require an extremely specialised knowledge and a high degree of chemical skill. This situation was mirrored within the area of peptide synthesis, until Merrifield et al proposed and developed Solid Phase Peptide Synthesis (SPPS) over thirty years ago (Merrifield, 1963). In SPPS immobilisation of the first amino acid of the required sequence to an insoluble resin enabled large excesses of reagents to be used to achieve the coupling of the second amino acid. Any unused materials remaining at the end of the coupling step could then be removed simply by washing the resin beads. This technology meant that the chemist could drive each coupling reaction to almost quantitative yields, and since the peptide intermediates formed were still bound to the resin, purification after each acylation step was not required. SPPS enables peptide and polypeptide synthesis to be employed as a routine research and synthetic tool, and permits large-scale combinatorial synthesis of peptides for screening of potential pharmaceutical agents.
For many years chemists have attempted to transpose this solid-phase methodology to oligosaccharide synthesis, with varying degrees of success. The first attempt was approximately 25 years ago (Frechet and Schuerch, 1971; Frechet and Schuerch, 1972; Guthrie et al, 1971; Guthrie et al, 1973). However, the ozone-mediated deprotection product was an aldehyde-substituted glycoside. Danishefsky and coworkers described the solid phase synthesis of the Lewis b Antigen (Randolph et al, 1995) and N-linked glycopeptides (Roberge et al, 1995) by initial attachment of the primary sugar unit of the oligosaccharide to a 1% divinylbenzene-styrene co-polymer support via a silyl ether linkage. The resin-bound sugar moeity was in this instance a glycal, with on-resin activation achieved via epoxidation of the double bond, and the resulting glycal residue acting as a sugar donor through nucleophile ring-opening of the epoxide. Since there are no calorimetric methods available to the sugar chemist to monitor on-resin glycosylations, the only means of assessing the progress of the reaction is by lysis of the oligosaccharide-resin bond and subsequent analysis of the cleavage product, usually by thin layer chromatography. The tetra-n-butylammonium fluoride-mediated deprotection conditions required to cleave Danishefsky""s silyl ether linker are both hazardous and slow. This coupled with the requirement for on-resin activation of the tethered glycals, makes the overall strategy and methodology far from ideal.
In an alternative approach, Douglas and coworkers described the synthesis of D-mannopentose using a polyethyleneglycol xcfx89-monomethylether co-polymer and a succinoyl or an xcex1,xcex1xe2x80x2-dioxyxylyl diether linker (Douglas et al, 1995). The reactions were carried out in solution phase, with removal of unused reactants being achieved by precipitation of the oligosaccharide-polymer complex and subsequent washing. In the latter example, cleavage of the oligosaccharide-polymer bond was achieved through catalytic hydrogenation, which required exposure of the conjugate to 1 atm of H2 for 48 h to achieve respectable yields. This again is far too slow to allow effective monitoring of individual glycosylation reactions. Yan et al reported sulphoxide-mediated glycosylation on a Merrifield resin, using a thiophenol linker for the attachment of the primary sugar residue (Yan et al, 1994). This method resulted in the construction of (1-6)-linked oligosaccharides, and was suitable for synthesis of both xcex1- and xcex2-glycosidic linkages. However, the thioglycosidic linkage to the resin dictates that similar sugar donors cannot be employed in this strategy.
Recently Rademann and Schmidt reported the use of trichloroacetimidate sugar donors to a resin bound sugar tethered via an alkyl thiol (Rademann and Schmidt, 1996); once again, however, this method precludes the use of the far superior thioglycoside sugar donors. Meanwhile, Adinolfi et al described the synthesis of disaccharides using a polyethyleneglycol-polystyrene resin, with connection of the first sugar to the polymeric support through a succinate spacer (Adinolfi et al, 1996). However, the acid lability displayed by this linker means that the primary sugar cannot be linked to the resin via the glycosidic position.
The above examples serve to illustrate that the critical element in solid phase synthesis is the nature of the linker between the solid support and the initial synthon. The linker must display excellent stability to the conditions of coupling and deprotection, yet in the case of solid phase oligosaccharide synthesis, it should also be rapidly and efficiently cleaved to allow monitoring of the progress of individual coupling reactions. The cleavage should ideally be achieved by the use of a relatively innocuous chemical reagent.
It is clear, then, that there remains a need in the art for simple, efficient and economical methods for solid-phase synthesis of oligosaccharides.
A hydrazine-labile primary amino-protecting group, N-1-(4,4-dimethyl-2,6-dioxocyclohexylidene)ethyl (Dde), has been reported for protection of lysine side chains during SPPS (Bycroft et al, 1993). This group was modified for use as a carboxy-protecting group in SPPS when the 2-(3-methylbutyryl)dimedone analogue of 2-acetyl-dimedone was condensed with 4-aminobenzylalcohol to afford 4-[N-[1-(4,4-dimethyl-2,6-dioxocyclohexylidene)-3-methyl-butyl]-amino]benzyl ester (ODmab)(Chan et al, 1995). 
The two protecting groups were reported to be stable to the deprotecting conditions widely used in SPPS, ie. trifluoroacetic acid (TFA) or 20% piperidine in dimethyl formamide (DMF). The ethyl ester, 4-[N-(1-(4,4-dimethyl-2,6-dioxocyclohexylidene)ethyl)amino]benzyl ester (ODab) showed small but significant instability to 20% piperidine-DMF. Both Dde and ODmab are linked to groups on amino acids, rather than directly to the solid-phase support. Their use in solid-phase oligosaccharide synthesis has not been suggested.
We have now surprisingly found that protecting groups similar to Dde and ODmab can be coupled to a polymeric support, thereby generating a system for the immobilisation of sugars. To this end we have immobilised N- and O-glycosides to the solid support and synthesised oligosaccharides using various sugar donors. The linkers display excellent stability to most acids and secondary/tertiary bases encountered in modern synthetic chemistry, yet are rapidly and efficiently cleaved with either ammonia, hydrazine or primary amines.
Bannwarth et al have independently developed a different solid phase linker around the Dde protecting group, which they have utilised for the immobilisation of amino acids and primary amines for combinatorial library synthesis (Bannwarth et al, 1996). However, the synthesis of this linker is both protracted and inefficient, and the linker only displays a limited stability to secondary bases such as piperidine. There has been no suggestion that this linker could be used for oligosaccharide synthesis. The linkers we have developed demonstrate a far greater stability than those of Bannwarth et al.
In one aspect, the invention provides a support for solid-phase synthesis of oligosaccharides, said support comprising:
a) a resin,
b) a linker covalently attached to the resin, and
c) one or more saccharide groups covalently attached to the resin via the linker,
wherein the linker is a 2-substituted-1,3-dioxocycloalkane compound, and
said support having general formula I: 
in which
R1 and R2 may be the same or different, and is each hydrogen or C1-4 alkyl;
Rxe2x80x2 is an amino sugar, a glycosylamine, or a glycosylamine of an oligosaccharide; a mono or oligosaccharide coupled through an alkyl-, substituted alkyl-, aryl-, substituted aryl-, cycloalkyl-, or substituted cycloalkyl-amino group; or a mono or oligosaccharide coupled through a carboxyalkyl-, substituted carboxyalkyl-, carboxyaryl-, substituted carboxyaryl-, carboxycycloalkyl-, or substituted carboxycycloalkyl-amino group; and
Rxe2x80x3 is an alkyl, substituted alkyl, aryl, substituted aryl, cycloalkyl, or substituted cycloalkyl spacer group which is directly coupled to the resin support, or which may optionally be coupled to the resin support via a suitable covalent linkage, which is stable to conditions of oligosaccharide synthesis and cleavage.
The covalent linkage to the resin may suitably be provided by a xe2x80x94CONHxe2x80x94, xe2x80x94Oxe2x80x94, xe2x80x94Sxe2x80x94, xe2x80x94COOxe2x80x94, xe2x80x94CHxe2x95x90Nxe2x80x94, xe2x80x94NHCONHxe2x80x94, xe2x80x94NHCSNH, or xe2x80x94NHNHxe2x80x94 grouping, eg. Spacer-CONH-resin, Spacer-O-resin, Spacer-S-resin, Spacer-CO2-resin, Spacer-CHxe2x95x90N-resin, Spacer-NHCONH-resin, Spacer-NHCSNH-resin, Spacer-NHNH-resin. Other possible covalent linking groups will be known to those skilled in the art.
Preferably both R1 and R2 are methyl.
Preferably Rxe2x80x2 is an oligosaccharide-Oxe2x80x94CH2xe2x80x94(C6H4)xe2x80x94NH, monosaccharide-Oxe2x80x94CH2xe2x80x94(C6H4)xe2x80x94NH, amino-oligosaccharide-CO2CH2xe2x80x94(C6H4)NH, or amino-monosaccharide-CO2CH2xe2x80x94(C6H4)xe2x80x94NH group.
In a particularly preferred embodiment the 2-substituted-1,3-dioxocycloalkane linker is functionalised Dde, Ddh or ODmab. In one very particularly preferred embodiment the support comprises a resin, a linker and a monosaccharide, an oligosaccharide, an aminosaccharide or an amino-oligosaccharide.
In a second aspect, the invention provides a support for solid-phase synthesis comprising a resin and a linker group, wherein the linker is a 2-substituted-1,3-dioxocycloalkane of general formula II: 
in which
X is OH or NH2; 
R1 and R2 may be the same or different, and is each hydrogen or C1-4 alkyl; preferably both R1 and R2 are methyl; and
Rxe2x80x3 is an alkyl, substituted alkyl, aryl, substituted aryl, cycloalkyl, or substituted cycloalkyl spacer group which is directly coupled to the resin support, or which may optionally be coupled to the resin support via a suitable covalent linkage, which is stable to conditions of oligosaccharide synthesis and cleavage. The covalent linkage may suitably be provided by a xe2x80x94CONHxe2x80x94, xe2x80x94Oxe2x80x94, xe2x80x94Sxe2x80x94, xe2x80x94COOxe2x80x94, xe2x80x94CHxe2x95x90Nxe2x80x94, xe2x80x94NHCONHxe2x80x94, xe2x80x94NHCSNH, or xe2x80x94NHNHxe2x80x94 grouping, eg. Spacer-CONH-resin, Spacer-O-resin, Spacer-S-resin, Spacer-CO2-resin, Spacer-CHxe2x95x90N-resin, Spacer-NHCONH-resin, Spacer-NHCSNH-resin, Spacer-NHNH-resin. Other possible covalent linking groups will be known to those skilled in the art.
In a third aspect, the invention provides a linker-saccharide complex, comprising a linker group of general formula II as defined above and a saccharide group as defined above for Rxe2x80x2.
In a fourth aspect the invention provides a linker compound carrying functional groups suitable to attach a primary amine to a resin via covalent bonds which are stable to conditions of oligosaccharide synthesis and cleavage, said compound having general formula III 
in which
X is OH or NH2;
R1 and R2 may be the same or different, and is each hydrogen or C1-4 alkyl, and
Rxe2x80x3 is an alkyl, substituted alkyl, aryl, substituted aryl, cycloalkyl, or substituted cycloalkyl spacer group, which carries a functionality capable of reacting with a functionalised resin.
Preferably the linker compound is 6-hydroxyl-6-(4,4-dimethyl-2,6-dioxocyclohexylidene)-hexanoic acid or an ester thereof. Preferably the ester is a benzyl, methyl, or t-butyl ester.
For the purposes of this specification the term xe2x80x9csubstitutedxe2x80x9d in the definitions of substituents within this specification means that the substituent is itself substituted with a group which does not change the general chemical characteristics of the substituent. Preferred such further substituents are halogen, nitro, amino, hydroxyl, and thiol; preferred halogens are chlorine and iodine. The person skilled in the art will be aware of other suitable substituents of similar size and charge characteristics which could be used as alternatives in a given situation.
For the purposes of this specification a compound is regarded as xe2x80x9cstable to conditions of oligosaccharide synthesis and cleavagexe2x80x9d if there is less than 10% loss of the compound after exposure at room temperature to ammonia, hydrazine or a primary amino compound in water or DMF. The person skilled in the art will readily be able to determine whether the stability of a particular compound is adequate for it to be useful for the purposes of the invention, using conditions appropriate for his or her particular needs.
The linker compound of the invention may be synthesized on the resin, or may be synthesized in solution.
The invention also provides kits useful in solid phase synthesis or combinatorial synthesis of oligosaccharides, comprising either
a) a resin-linker-saccharide support,
b) a linker-saccharide complex, or
c) a resin-linker support,
according to the invention, as described above. The kit may optionally also comprise one or more further reagents such as protecting agents, deprotecting agents, and/or solvents suitable for solid phase or combinatorial synthesis. The person skilled in the art will be aware of suitable further reagents. Different types of kit can then be chosen according to the desired use.
The resin may be any resin which swells in water and/or in an organic solvent, and which comprises one of the following substituents: halogen, hydroxy, carboxyl, SH, NH2, formyl, SO2NH2, or NHNH2, for example methylbenzhydrylamine (MBHA) resin, amino or carboxy tentagel resins, 4-sulphamylbenzyl AM resin. Other suitable resins will be known to those skilled in the art.
The invention also provides a method of solid-phase synthesis of oligosaccharides, comprising the step of sequentially linking mono- or oligosaccharide groups to a support as described above. Similarly the mono- or oligosaccharide building blocks may be as described above.
This method is particularly useful for combinatorial synthetic application.
The linker compound may be synthesised in solution or directly on the resin in a stepwise manner prior to the coupling of the initial sugar group, or the linker-initial sugar conjugate may be synthesised in solution phase and subsequently coupled to the solid support, with subsequent sugars being sequentially attached. Preferably the second and all subsequent sugar groups are coupled to the oligosaccharide chain-resin conjugate after the last sugar in the oligosaccharide chain is partially deprotected.
The invention accordingly provides a method of synthesis of a linker group according to general formula I as defined above, comprising the step of C-acylation of a 2-substituted 1,3-dioxocyclohexane compound with a dicarboxylic acid. Preferably the dicarboxylic acid is mono-protected by ester formation. More preferably the reaction is activated with carbodiimide and catalysed by N,Nxe2x80x2-dimethylaminopyridine.
The product of the reaction may optionally be reacted with 4-aminobenzyl alcohol, to form the 4-aminobenzyl derivative.
The invention also provides a method of synthesis of a resin-linker support, comprising the step of swelling a resin in a suitable solvent, treating the swollen resin with a dicarboxylic acid, and reacting the thus-produced product with a 2-substituted 1,3-dioxocycloalkane compound. Preferably for both synthesis of the linker and synthesis of the resin-linker support the 2-substituted 1,3-dioxocyclolkane compound is 5,5-dimethyl-1,3-cyclohexanedione. Also preferably the dicarboxylic acid is adipic acid.
The first sugars attached to the resin-linker unit may be unprotected, partially protected or fully protected glycosides, aminoglycosides, or ether- or amino-linked sugars, where the coupling takes place through a non-glycosidic position.
The building block mono- or oligosaccharide-donors may be any activated sugar, including but not limited to orthoesters, thioorthoesters, cyanoalkylidene derivatives, 1-O-acyl sugars, amino sugars, acetimidates, trichloroacetimidates, thioglycosides, aminoglycosides, amino-oligosaccharides, glycosylamines of oligosaccharides, glycosyl thiocyanates, pentenyl glycosides, pentenoylglycosides, isoprenyl glycosides, glycals, tetramethylphosphoro diamidates, sugar diazirines, selenoglycosides, phosphorodithioates, glycosyl-dialkylphosphites, glycosylsulphoxides and glycosylfluorides.
Preferably the first sugar coupled to the resin is an aminosugar, an aminoglycoside, or an amino-oligosaccharide or a glycosyl amine of an oligosaccharide.
Preferably partial sugar deprotection is achieved by using acyl-type, trityl, benzyl-type, acetal-type, or various silyl and/or photolabile protecting groups in addition to permanent protecting groups. This permits the synthesis of branched oligosaccharides by using two orthogonal hydroxy-protecting groups on a single sugar donor.
The synthesised oligosaccharide can be cleaved from the resin using ammonia, hydrazine or a primary amine, such as butylamine or cyclohexylamine. For the preparation of aminoglycosides, ammonia or a suitable primary amine in an organic solvent is preferably employed. For the preparation of hydrazides, hydrazine in water or in an organic solvent is preferably employed. For the preparation of oligosaccharides, ammonia in water or in an organic solvent is preferably employed, followed by acidification. When the linker contains a 4-aminobenzyl moiety, after cleavage as described above the first sugar is released still protected by the aminobenzyl group; this can be removed by hydrogenation if desired.
The person skilled in the art will appreciate that the oligosaccharide can be retained on the resin for use as an analytical or preparative reagent, for example in affinity chromatography or for bulk-scale affinity separation.